In article ,
Richard J Kinch wrote:
Norman Yarvin writes:

So suppose you
have two containers, one filled with gas, and another empty (filled
with vacuum). You connect them, and open the valve between them.
The gas molecules bounce around into the newly connected container,
so now you have two containers each with half the gas. But the gas
temperature is the same (or very close).

No, adiabatic expansion, temperature falls.

That's what it'd be if you expanded the gas against a piston. Letting
the gas pass through a valve is different. The energy that would go
into pushing the piston instead goes into reheating the gas. The net
result is that the temperature of the gas does not change.

I would expect that the two containers would end at different temperatures,
and thus at least one must have changed. Where the heat from the work of
expansion ends up depends on several things, including the geometry of the
valve and connections.

Yes, it's the average temperature -- the total amount of energy in the
gas -- that will be unchanged. Generally the tank providing the gas will
cool, and the other one will heat. Now you could convert some of that
temperature difference into mechanical work, by running a Carnot engine
between the two containers, but you still wouldn't get as much work from
it as you would have done by just expanding the gas against a piston.

On a microscopic level, what happens with a valve or a regulator is that
the walls are motionless, and bounce gas molecules back at the same speed
as they came in with. A piston moving away, on the other hand, bounces
gas molecules back more slowly than they came in, since each collision is
with a retreating surface; this cools the gas.


--
Norman Yarvin http://yarchive.net